AMD & Cohere broaden AI model partnership for global clients
AMD and Cohere have expanded their global collaboration to offer enterprise AI models that run on AMD Instinct GPUs and to integrate Cohere's North platform into AMD's own internal AI systems.
The collaboration gives Cohere customers the ability to deploy AI offerings, including Command A, Command A Vision, Command A Translate, and North, using AMD Instinct GPU-powered infrastructure. This development is aimed at supporting enterprise and sovereign AI initiatives globally, with particular attention to meeting performance goals and optimising total cost of ownership (TCO).
New options for enterprise AI
The partnership enables a broader set of options for organisations and government entities seeking scalable and secure AI solutions. By leveraging AMD Instinct GPUs for Cohere's models, enterprises can choose how to deploy AI technologies across different environments, including projects that require strict data, security, and compliance controls.
Nick Frosst, Co-Founder at Cohere, said:
"We are excited to further expand our relationship with AMD with the availability of many of our models and North on AMD Instinct GPUs. Given public and private sector customers the ability to deploy our full suite of technology across our foundational models and security-optimized enterprise AI products gives them significantly greater flexibility in how they choose to deploy Cohere's AI. AMD has a compelling TCO proposition with its AI infrastructure and is a great option for sovereign AI initiatives in Canada and globally."
With this collaboration, customers using Cohere's enterprise AI products can deploy these solutions with higher performance and efficiency, thanks to the high memory capacity and bandwidth of AMD's AI infrastructure. This setup reduces the needed deployment footprint and allows organisations to manage costs more effectively.
Cohere models on AMD AI
AMD will also incorporate Cohere's North platform into its internal enterprise AI portfolio, making the technology a core component for AMD's engineering and business AI workloads. This adoption signifies the company's intention to both use and demonstrate the capabilities of third-party AI solutions within its operations.
Vamsi Boppana, Senior Vice President, AI at AMD, commented on the significance of the expanded relationship:
"Cohere's full-stack AI solutions are now deployable on AMD Instinct infrastructure, empowering organizations and governments to scale AI with exceptional performance, efficiency, and memory capacity. With the AMD AI computing platform, customers benefit from excellent total cost of ownership and energy efficiency - critical advantages as enterprise and governments accelerate their AI transformation."
The companies have already been engaged in optimising Cohere's large language models (LLMs) on AMD Instinct GPU platforms. According to both parties, these models have been implemented across various enterprises and are delivering the required performance for complex workloads involving long-context reasoning and large datasets.
Serving sovereign and enterprise needs
With the need for secure, private, and compliant AI models growing, especially within government and regulated sectors, the ability to deploy Cohere's enterprise models on AMD's infrastructure is presented as beneficial. Both companies state this will allow sovereign AI deployments that adhere to national data handling and security requirements, providing a foundation for trusted AI systems that organisations and governments can control directly.
The collaboration between AMD and Cohere follows ongoing work to expand AI infrastructure capabilities in Canada and globally, supporting the development of systems that can operate within the legal and operational frameworks of multiple jurisdictions. This structure provides customers the scope to determine how and where their AI initiatives are deployed, with attention to their unique needs regarding performance, cost, and compliance.
